Anyscale
Anyscale is an AI model provider.Tokenando tracks 4 Anyscale models, with input pricing from $0.150/M and an average blended cost of $0.662/M. Its flagship model is Llama 3.1 70B (Anyscale).
Ray-native managed inference and training. Strongest fit when distributed Python workloads sit alongside model serving.
MODELS TRACKED
4
1 category
FLAGSHIP
Llama 3.1 70B (Anyscale)
Live API
MIN INPUT
$0.150/M
cheapest model in family
AVG BLENDED
$0.662/M
across 4 priced models
MAX CONTEXT
128,000
largest window in family
Efficient
4 models
Llama 3.1 70B (Anyscale)profile
Live API · 128,000 ctx
in $1.000/Mout $1.000/M
Ray-based inference · manual-seed
Mistral 7B (Anyscale)profile
Live API · 32,000 ctx
in $0.150/Mout $0.150/M
Anyscale serverless · manual-seed
CodeLlama 70B (Anyscale)profile
Live API · 16,000 ctx
in $1.000/Mout $1.000/M
Code-tuned Llama · manual-seed
Mixtral 8x7B (Anyscale)profile
Live API · 32,000 ctx
in $0.500/Mout $0.500/M
MoE on Anyscale · manual-seed
Frequently Asked Questions
How many models does Anyscale offer?
Tokenando tracks 4 Anyscale models.
How much do Anyscale models cost?
Anyscale model input pricing starts at $0.150 per million tokens, with an average blended cost of $0.662 per million across the 4 priced models we track.
What is Anyscale's flagship model?
Anyscale's flagship model is Llama 3.1 70B (Anyscale). It is the highest-tier Anyscale model we track, with input pricing of $1.000 per million tokens.
What model categories does Anyscale cover?
Anyscale covers 1 category: efficient.